Given data on San Francisco's real estate development pipeline, I've been tasked with finding where multifamily projects are being proposed and approved around the city, in order to inform a West Coast investment and development strategy. I'm told by my client, a real estate startup founding partner, that knowing where activity is prevalent will help them narrow their search for opportunities, as it will allow them to begin to understand whether a submarket is over- or under-saturated. Likewise, insights on approval frequency will also inform an understanding of what projects are at risk of being disapproved or delayed by neighborhood-specific contingents that can block projects.
The data itself is posted every quarter by the city of S.F. and includes details on all active project and permit files for all types of real estate applications, including non-multifamily ones (i.e. single-family, office, retail, etc.) and applications that have been relatively dormant (e.g. included every quarter but, for whatever reason, not moving through the approval process from application to construction.) I worked with quarterly files since 2015, filtering each to only multifamily (2 or more units) projects and those having filing statuses new to a given quarter to limit duplicates.
To start, I gauge market saturation by seeing how new construction is spread across S.F.'s 41 neighborhoods.
Data Source: DataSF. Counting construction starts.
We see that the combined area of FiDi/South Beach and south of Market Street – and particularly, the SoMa neighborhood itself – is where the greatest concentration of recent new construction has occurred. Leading the way is SoMa, which has seen more than 13,000 units start construction.
I then look to see how effective approval processes may differ around the city by breaking new filing statuses down by supervisor district, rather than neighborhood. I also choose to see how counts compare when we control for land area to more fairly compare filing and approval tendencies across districts. Below, it shouldn't come as a surprise to see that Districts 6 and 9, which hold FiDi and SoMa, respectively, have seen the most new planning filings. After adjusting for differences in size, though, we see that District 3 (and 8, to a lesser degree) has seen a relatively large share of activity. This could be because it holds part of FiDi despite being the smallest district.
Data Source: DataSF. Counting new planning applications, approvals
The above also shows how often approvals come relative to filings. For a better idea of approval tendencies across districts, I finish by mapping ratios of approvals-to-applications.
Data Source: DataSF. Dividing number of planning approvals by applications
We see that District 10, which includes the Bayview Hunters Point and Portrero Hill neighborhoods and part of Mission Bay, has seen the highest ratio of approvals to applications. Distant runners-up are Districts 6 and 9, which include the Mission and SoMa neighborhoods.
These and earlier findings indicate the Bayview Hunters Point and Potrero Hill neighborhoods might offer the most attractive combo of higher growth trajectory and lower disapproval risk. Though recently lower growth compared to SoMa and FiDi, the District 10 neighborhoods' proximity to higher growth areas and their own higher approval rate suggest they offer a safer investment, and one that could benefit from being near SoMa and FiDi.
For these reasons, I recommend that the firm's current property-selection efforts zero in on District 10 sites but that it hold off on any action until completing further research. More analysis is needed to understand whether the area's relative under-saturation is a flaw or a feature. Furthermore, given the time element inherent to the approval process, the above ratio might be capturing an uptick in applications ocurring before this analysis' scope leading to more current approvals, as opposed to only approval-friendliness. However, the gross and relative number of approvals in the area still warrant serious consideration.